FuzzyFriend | Devpost

Homepage
Meet the Friends
About
Friend description
Chat interface

Inspiration

Hearing a Huawei representative discuss the issues of mental health resonated strongly with our group. Younger generations are increasingly affected by mental health issues, especially as many of us experience the impacts of isolation and loneliness due to COVID-19. We were also inspired by the challenge of creating something ‘warm and fuzzy’. We wanted to create a fun and lighthearted solution that attempts to combat some of these issues by providing one with a friend for whatever mood they are in.

What it does

This website provides users with a “fuzzy friend”, a chatbot tailored to provide feel-good suggestions based on the user’s emotions. Users are introduced to 6 fuzzy friends, each with their own personalities, which they can then select based on their mood. Different bots offer various responses to messages, whether links to videos, jokes, or friendly commentary.

How we built it

We built this project using Python’s ChatterBot and Flask frameworks, and HTML. We worked on both front-end and back-end development, using Bootstrap to help with the front-end. We also attempted to use Heroku as a server to link the two components.

Challenges we ran into

Our main struggle was connecting our website to the server, so that users without Flask or ChatterBot could still use our FuzzyFriend chat service. We were new to GitHub and struggled with merging our front and back-end components without creating conflict. We also struggled to download and import the libraries necessary for our website to work, such as Flask. Building an AI chatbot capable of improving itself is also something that is harder than it seems!

Accomplishments that we're proud of

We managed to build a clean, user-friendly looking website that matches our warm, friendly theme. Another accomplishment was working with new libraries and improving our technical skills, such as learning to use Flask and how to collaborate on github. As our first hackathon ever, and first collaborative project, learning how to use github and creating common code that would work on all of our computers is an accomplishment that we are proud of. There were a lot of new components to this project, but we were all able to adapt and work with them to create what we wanted. Finally, we are proud of our teamwork. We were all incredibly supportive, determined and helpful through each step in the process; we were organized in planning our project, assigning and delegating different tasks, as well as quick to help eachother out.

What we learned

We learned how to collaborate on GitHub, how to implement libraries and frameworks we were not familiar with, and how to have fun coding together!

What's next for FuzzyFriend

We want to make our chatbots smarter to make user conversations more fluid. Currently, our bots are limited to very restricted conversations and often misunderstand messages. In the future we want to train our robots to interact more appropriately with all sorts of prompts from the user while also staying true to their character. We would also like to successfully connect our product to a server and domain. This way, users could more easily access its contents.